Projects involving data centers in Earth orbit are generating the prospect of a billion-dollar market for the insurance sector. Companies such as SpaceX and Blue Origin are developing initiatives to position data processing centers in space, while insurance companies analyze how to cover infrastructure that lacks sufficient risk history for accurate calculations.
These proposals, anticipated for 2026, include the use of constellations composed of thousands of satellites, aiming to concentrate computational power outside of Earth. Google is also investigating a similar architecture, and the startup Starcloud has already demonstrated capability by placing an Nvidia graphics processor in orbit.
The potential migration of part of the artificial intelligence infrastructure to space inaugurates a new avenue for the insurance market, but simultaneously introduces significant challenges in terms of financial capacity, regulation, pricing definition, and technical risks, which may require new launches for repairs or replacements.
The boldest initiative belongs to SpaceX. In January, the company submitted an application to the U.S. Federal Communications Commission regarding a constellation that could reach one million satellites. The goal is to use this structure to configure an orbital center focused on artificial intelligence computing.
Elon Musk argues that the combination of solar energy available in space, decreasing launch costs, and increased spending to expand energy supply in terrestrial facilities could make this model economically viable. According to his assessment, this transition could occur within a period of two to three years.
Blue Origin, Jeff Bezos's company, is taking a similar route, although with a timeline perceived as more extensive. In March, the company presented plans for 51,600 satellites intended for low Earth orbit data centers. Bezos considered the idea feasible but found the two-to-three-year timeframe excessively optimistic.
This movement is not limited to the two mentioned corporations. Google is researching a project called Project Suncatcher, which consists of a network of interconnected satellites powered by solar energy and equipped with their own artificial intelligence chips. Meanwhile, Starcloud has already executed a practical test by sending an Nvidia H100 unit into space.
If these endeavors reach the projected scale, the value of the equipment installed in space could amount to hundreds of billions of dollars. It is in this context that an opportunity arises for insurers, who would face a new class of extremely high-value assets.
Patton Kline, head of aviation and space at Marsh in the United States, emphasizes that the insurance segment must keep pace with this change. His analysis indicates that companies focused exclusively on covering terrestrial assets run the risk of missing a growth opportunity if orbital computing becomes established.
Currently, it is estimated that about 30 insurers globally specialize in space coverage. The market moves between US$ 500 million and US$ 750 million in annual premiums, a modest amount compared to the size needed to protect an orbital infrastructure estimated at hundreds of billions of dollars.
Expansion can also offer an attraction to the sector: risks associated with space assets would not be directly linked to events that typically affect ground structures, such as earthquakes and hurricanes. However, increasing coverage capacity for this type of infrastructure will not be a simple task.
Andreas Berger, CEO of the reinsurance group Swiss Re, observes that orbital computing merges two rapidly expanding areas: artificial intelligence infrastructure and the commercial space sector. However, this convergence creates gaps that complicate the creation of sustainable insurance products.
Among the obstacles are the lack of sufficient references to measure the frequency and impact of certain events, as well as uncertainties regarding regulation and the availability of capital to absorb large losses. For insurers, the primary challenge lies in pricing without relying on sufficiently reliable models.
Technical problems exacerbate this uncertainty. Components in orbit are exposed to failures during launch, radiation, hardware defects, thermal control difficulties, and collisions caused by the increase in space debris.
There is also a crucial distinction compared to terrestrial data centers: a failure in space equipment cannot simply be resolved by the arrival of a maintenance team. Depending on the nature of the problem, a new launch will be necessary to perform the repair or replacement of the affected item.
Thus, the insurance sector finds itself in a complex position. On one hand, the expansion of orbital data centers can generate a multi-billion dollar coverage market; on the other hand, the absence of established regulation, adequate capital, and models capable of quantifying risks makes the speed at which this market will mature uncertain.
For orbital computing projects to advance, it is not enough just to put the equipment into orbit. It is imperative to also establish conditions so that investors, space companies, and insurers can estimate losses, define values, and determine who will assume the risks of an infrastructure that is still unprecedented on a commercial scale.
